You are on page 1of 5

Review

Epidemiological features of chronic low-back pain

Gunnar B J Andersson

Although the literature is filled with information about the prevalence and incidence of back pain in general, there is
less information about chronic back pain, partly because of a lack of agreement about definition. Chronic back pain
is sometimes defined as back pain that lasts for longer than 712 weeks. Others define it as pain that lasts beyond
the expected period of healing, and acknowledge that chronic pain may not have well-defined underlying pathological
causes. Others classify frequently recurring back pain as chronic pain since it intermittently affects an individual over
a long period. Most national insurance and industrial sources of data include only those individuals in whom
symptoms result in loss of days at work or other disability. Thus, even less is known about the epidemiology of
chronic low-back pain with no associated work disability or compensation. Chronic low-back pain has also become
a diagnosis of convenience for many perople who are actually disabled for socioeconomic, work-related, or
psychological reasons. In fact, some people argue that chronic disability in back pain is primarily related to a
psychosocial dysfunction. Because the validity and reliability of some of the existing data are uncertain, caution is
needed in an assessment of the information on this type of pain.

Back pain in society Study Lifetime Prevalence % Study group


The prevalence of chronic back pain should be placed in incidence
Point Period Number Age Sex
the context of the prevalence of back pain in general. (years) (M/F)
Many studies attest to the high frequency of back Biering-Sorensen7 626 120 449 3060 M
complaints in society. 7085% of all people have back Biering-Sorensen7 614 152 479 3060 F
pain at some time in life. The annual prevalence of back Frymoyer, et al8 699 1221 2855 M
Gyntelberg9 25 4059 M
pain ranges from 15% to 45%, with point prevalences Hirsch, et al10 488 692 1572 F
averaging 30%.1 In the USA, back pain is the most Hult11 600 1193 2559 M
common cause of activity limitation in people younger Magora12 129 3316 M, F
Nagi, et al13 180 1135 1864 M, F
than 45 years, the second most frequent reason for visits Papageorgiou, et al14 590 35 1884 >18 M
to the physician, the fifth-ranking cause of admission to Papageorgiou, et al14 590 42 2617 >18 F
hospital, and the third most common cause of surgical Svensson, et al15 61 31 716 4047 M
Svensson, et al16 67 35 1640 3864 F
procedures.24 About 2% of the US workforce are Valkenburg, et al17 514 222 3091 >20 M
compensated for back injuries each year. Valkenburg, et al17 578 302 3493 >20 F
Data from other western countries are similar. UK Walsh, et al18 583 36 2667 2059 M,F
estimates place low-back pain as the largest single cause Data from Andersson.1
of absence from work in 198889, and it is responsible Table 1: Prevalence and lifetime incidence of low-back pain in
for about 125% of all sick days.5 This figure is similar to cross-sectional studies
data from Sweden where, since 1961, 1119% of all
annual sickness absence days are taken by people with a and spine impairments the most frequently reported
diagnosis of back pain.1,6 In 1987, 148 million workdays subcategory of musculoskeletal impairment (517%). The
were lost in Sweden because of back pain, which annual rates varied significantly by sex and age (table 2).
constitutes about 135% of all reported sick days. Overall, Back and spine impairments were more common in
8% of the insured Swedish population were listed as sick women (703 per 1000 population) than in men (573
with a diagnosis of back pain at some time during 1987. per 1000 population), and more common among white
Cross-sectional surveys of local populations people (687 per 1000 people) than black people (387
corroborate the data from the national surveys. per 1000 people). In 1988, back and spine impairments
Prevalence rates of back pain are difficult to compare resulted in over 185 million days of restricted activity
because of the time of the sampling, the sampling (210 per impairment), which included 83 million days
technique, and the actual questions asked. Representative confined to bed (54 per impairment; table 3). About
data that range from 12% to 35% are shown in table 1. 56% of days of restricted activity occurred among
women. Rossignol and colleagues19 followed a cohort of
Chronic back pain 2341 cases that were randomly sampled to represent
Praemer and colleagues2 used the 1988 National Health individuals who had been compensated for occupational
Interview Survey (198588) to estimate the frequency of back injury in the state of Quebec, Canada, in 1981.
chronic or permanent impairment in the USA. 67% of the sample were still absent from work after 6
Musculoskeletal impairment was the most prevalent months, which accounted for 68% of work days lost and
impairment in people aged up to 65 years, and back 76% of the total compensation cost for low-back pain.
When the cumulative absence was calculated over 3
Lancet 1999; 354: 58185 years, 97% of workers were absent for 6 months or
Department of Orthopedic Surgery, Rush-Presbyterian-St Lukes longer, which illustrates the recurrent nature of back
Medical Center, Chicago, IL 60612, USA (G B J Andersson MD) pain. A logistic-regression model was used to calculate

THE LANCET Vol 354 August 14, 1999 581


Number in millions Prevalence per 1000 population Study Time (weeks) to return to work
Total impairments 15 431 641 1 2 4 6 8 12 24 52
Sex Andersson, et al37 46 63 75 88 94 97 98
Male 6701 574 Choler, et al38 60 67 88 95 98 99
Female 8730 703 Spitzer, et al22 74 92 93 96
Webster and Snook39 42 62 79 87 89 93
Ethnic origin
van Doorn20 53 65 71 80 88 93
White 13 957 687
Black 1137 387 All data are % of population.
Other 336 Table 4: Proportion of population listed as sick with diagnosis
Age (years) of back pain by time taken to return to work
017 714 112
1844 8295 805 The possible aetiological importance of psychiatric
4564 4105 901 disorders was investigated by Polatin and colleagues21 in a
6574 1333 759
7584 780 872 group of 200 patients with chronic low-back pain who
.84 203 936 were entering a functional-restoration programme. The
Data from Praemer and colleagues.8 patients were assessed for current and lifetime psychiatric
syndromes; 77% of patients met lifetime diagnostic
Table 2: Prevalence of back and spine impairments in
criteria and 59% showed current symptoms for at least
USA in 1988
one psychiatric diagnosis, most commonly depression,
substance misuse, and anxiety disorders. In addition,
risk factors associated with absences from work of
51% met criteria for at least one personality disorder.
6 months or longer, and showed that age and site
These prevalence rates were significantly greater than
of symptoms were the two most important variables. A
rates in the general population. 54% of patients with
23-year increase in age doubled the odds of accumulating
depression, 94% of those with substance abuse, and 95%
at least 6 months of absence, and lumbar symptoms were
of those with anxiety disorders had experienced these
286 times more likely than thoracic symptoms to become
syndromes before the onset of their back pain. This
chronic. The odds ratios for sex and occupation were not
finding suggests that substance abuse and anxiety
significant. van Doorn20 reported that in a self-employed
disorders precede chronic low-back pain, whereas
subset of Dutch dentists, veterinarians, physicians, and
depression may develop before or after the onset of this
physical therapists, 23% of claims lasted longer than 6
type of back pain.
months or were deemed chronic. The risk of chronic back
pain increased with older age. With a predictive Coxs
regression model, van Doorn found that a specific Recovery from back pain
aetiological diagnosis, older age, previous back pain, and Most patients with back pain recover quickly and without
psychosocial disorders were factors that had a negative residual functional loss. Table 4 shows the similarities in
effect on recovery. recovery rates between studies and countries. Overall,
6070% recover by 6 weeks, 8090% by 12 weeks
Psychological distress and psychiatric (figure 1). Recovery after 12 weeks is slow and uncertain.
Fewer than half of those individuals disabled for longer
disorders
than 6 months return to work and, after 2 years of
Various cross-sectional studies indicate an association absence from work, the return-to-work rate is close to
between psychological factors and the occurrence of low- zero.22 Diagnosis has been found to affect recovery;
back pain.1 These factors include anxiety, depression, patients with sciatica recover more slowly than those with
somatisation symptoms, stressful responsibility, job back pain alone. Andersson and colleagues23 report that
dissatisfaction, mental stress at work, negative body 60% of patients with low-back pain recovered in 10 days,
image, weakness in ego functioning, and poor drive compared with 40% of patients with sciatica.23
satisfaction. The experience of stress, anxiety, and Compensation has a negative influence on the length
depression is sometimes, but not always, secondary to of disability. Sander and Meyers23 compared disability
back pain. In a few prospective studies, various periods for on-duty injuries and off-duty injuries.
symptoms that indicate psychological distress predicted The study included both lumbar sprains and strains and
the development of back disorders in people who did not patients who underwent an operation. The average time
have previous back pain. off work for an on-duty strain injury was 149 months,
compared with 36 months for off-duty injuries; the
Restricted activity Bed days per corresponding periods of absence for patients who had an
days per impairment impairment operation were 93 and 44. These data are supported by
Total number 120 54 Greenough and colleagues findings.25 In a retrospective
Sex cohort study of 300 patients in Adelaide, Australia, the
Male 145 59 average time off work for men with compensation was 12
Female 101 50 (range 02584) months versus 025 (0180) months for
Ethnic origin men with no compensation, for women, the
White 101 42 corresponding off-work periods were 15 (0132) and 05
Black 318 154
(022) months. The type of accident, diagnosis, and
Age (years)
>65 115 54
specific treatment did not affect outcome, whereas
65 150 53 psychological disturbance was a major determinant
Data from Praemer and colleagues.2 of outcome. Leavitt26 found that work-related back
Table 3: Restricted activity days and days confined to bed symptoms resulted in longer absence from work than
(bed days) for back or spine disorders in 1988 by major non-work-related symptoms, even after control for the
population subgroup influence of the physical work environment.

582 THE LANCET Vol 354 August 14, 1999


Recurrence
The recurrence rate of low-back pain is so high that it
seems to be part of its natural history. Table 5 shows data
from different studies. Lifetime recurrences of up to 85%
were reported by Valkenburg and Haanen,17 whereas in
Sweden, the 1-year recurrence of sick-listing for low-back
pain was 44% in 1987.6 Data from Canada show
recurrence rates of 20% in 1 year and 36% over
3 years.19,36 Men had a higher risk of recurrence than
women, and people aged 2544 years had the highest rate
of recurrence. Occupation also affected the rate of
recurrence, the highest rate occurred in nurses and
drivers and the lowest among white-collar workers.

Office visits to physicians


Figure 1: Time course of acute low-back pain In 1990, there were about 15 million office visits to
About 90% resolve in 3 months. Data from Andersson and colleagues.37 physicians in the USA for mechanical low-back pain,
which accounts for about 28% of all office visits.4 Since
Age has also been found to have a negative effect on Hart and colleagues4 did not include visits to allied health
recovery.19,20,2731 It is unclear whether the effect of age is professionals, such as chiropractors, the actual number of
due to a reduced ability to recover from an injury with office visits was probably more than 30 million per year.
older age or to socioeconomic or general health factors. Among physicians and osteopaths, the number of visits in
Psychological factors are known covariants of chronic 198990 was only slightly higher than in 198081 (when
low-back pain and of chronic pain in general. Their role it was 122 million). Another US study included
as predictors of recovery from back pain is unclear. chiropractors and divided the office visits into episodes
Frymoyer and Cats-Baril32 reported that psychological of spine care, which were defined as all visits believed to
variables at the beginning of a back pain episode did be associated with a particular episode of back pain; only
not predict long-term disability, whereas van Doorn20 people aged up to 65 were included.37 According to this
reported opposite findings. definition, there were 91 episodes per 100 person-years.
European statistics are similar. In 199293, there were 7
Predictive models of return to work million visits to general practitioners for back pain in the
There have been various attempts to develop models to UK. The annual rate of consultations with these patients
predict the duration of absence from work due to low- in the UK has been estimated at 5575% of the adult
back pain. These models are difficult to compare because population. Swedish estimates from 1987 for the number
of differences in the population studied, time of the of outpatient visits for back pain are 790 000 people or
evaluation, and socioeconomic differences between 89 per 100 person-years.
countries. Rossignol and colleagues19 used a logistic- With respect to chronic back pain, Kochs analysis44
regression model to analyse factors that influence the of the 198081 data discussed above, reported that
risk of work-related sickness absence for longer than 179% of all visits for chronic pain were for chronic back
6 months. Age and location of symptoms were the most pain. In about 47% of those visits, an analgesic was
important variables. Frymoyer and Cats-Baril 33 assembled prescribed. Data from other studies cited by Hart4 show
a panel of experts to reach consensus on the factors that that on an annual basis 41% of patients with chronic back
predict disabling low-back pain, and to assign weighting pain visit a physician, 30% a chiropractor, and 8%
to these factors. This expert model included factors on another type of care provider. Since patients visits to
injury, self-efficacy, demographic characteristics, history chiropractors average a higher number per case, Hart
of pain, and employment. The model was tested on a estimated that in a typical office, over half of the workers
group of patients from a low-back pain clinic, and visits for chronic low-back pain are to a chiropractor.
compared with an empirical model, developed by a
Study % of study population Time (years) Type of study population
logistic statistical technique. The work environment,
psychosocial factors, and the duration of the current Abenhaim35 200 1 Prospective
Occupational BP
episode of low-back pain were the main predictors. 363 3 Prospective
Other factors associated with long-term disability were Occupational BP
perception of fault, involvement of lawyers, self- Anderson40 89 2 Dockyard workers
Bergquist-Ullman41 220 1 Prospective
prediction of disability, income, educational attainment, Occupational BP
and employer attitudes. Physical and pure psychological Biering-Sorensen 42 38 (men) 1 Prospective
factors had no predictive value. Burton and colleagues34 39 (women) 1 Random sample
Choler, et al38 120 15 Work absence
used multiple-regression analyses and concluded that Perospective
persistent disability (after 1 year) was mainly related Moens, et al43 720 Lifetime Female family care
to psychosocial factors. In another study, Lancourt and employed
Nachemson6 440 1 Sickness absence data
Kettelhut35 concluded that non-organic factors were van Doorn20 86 1 Claims, self-employed
better predictors of return to work than organic factors. 160 2
Specifically, they reported that for a group of patients 200 3
470 8
who did not return to work within 6 months, previous
injuries and stability of family living arrangements were Table 5: Recurrence of back pain
important factors. BP=back pain.

THE LANCET Vol 354 August 14, 1999 583


large and costly; these individuals are also a cause of
major disability and absence from work. Back and spine
impairments are the most common impairment among
young and middle-aged people. Although back pain
seems to be equally common in men and women, back
and spine impairments are more common in women than
in men. Disability trends indicate large increases in all
developed countries. Among the factors that contribute to
long-term disability are: age, location of symptoms, and
legal, socioeconomic, and psychological factors. Rates of
surgery for back pain have increased rapidly over the past
15 years, whereas rates of non-surgical hospital admission
have decreased. Compared with other developed
countries, the surgical rates in the USA are high, but
there are also large variations between different US regions.
To reduce chronicity, disability, and cost, preventive
measures should be explored. Unfortunately, attempts to
prevent the occurrence of back pain (primary prevention)
have been unsuccessful and prevention of the negative
consequences of a back-pain episode may be more fruitful.
Figure 2: Ratios of rates of surgery for back pain in 11 countries
and Canadian provinces compared with rates in USA References
Data from Cherkin and colleagues. 46 1 Andersson GBJ. The epidemiology of spinal disorders.
In: Frymoyer JW, ed. The adult spine: principles and practice, 2nd ed.
Philadelphia: Lippincott-Raven, 1997: 93141.
Rates of hospital admission and surgery 2 Praemer A, Furnes S, Rice DP. Musculoskeletal conditions in the
Volinn and colleagues45 examined the US National United States. Rosemont: AAUS, 1992: 199.
3 Taylor VM, Deyo RA, Cherkin DC, Kreuter W. Low-back pain
Hospital Discharge Survey for time trends (197987), hospitalization: recent United States trends and regional variations.
and Taylor and colleagues3 extended the study to include Spin 1994; 19: 120713.
data up through 1990. During the 11 years surveyed, 4 Hart LG, Deyo RA, Cherkin DC. Physician office visits for low back
operations among adults for low-back pain increased by pain. Spine 1995; 20: 1119.
5 Frank A. Low back pain. BMJ 1993; 306: 90108.
55% from 147 500 in 1979 to 279 000 in 1990. This rise
6 Nachemson AL. Back pain. Causes, diagnosis and treatment.
corresponds to an increase from 102 to 158 per 100 000 Stockholm: The Swedish Council of Technology Assessment in
adults. This increase was particularly great for fusions Health Care, 1991.
that increased by 100% from 13 to 26 per 100 000 adults. 7 Biering-Sorensen F. Low back trouble in a general population of 30-,
30-, 50-, and 60-year old men and women. Study design,
In 1990, there were 46 500 lumbar fusions and 232 500 representativeness and basic results. Dan Med Bull 1989; 29: 28999.
operations without fusion for low-back pain. Non- 8 Frymoyer JW, Pope MH, Clements JH, Wilder DG, McPherson B,
surgical admissions to hospital, on the other hand, Ashikaga T. Risk factors in low back pain: an epidemiological survey.
decreased from 402 per 100 000 adults in 1979 to 150 J Bone and Joint Surg 1983; 65A: 213.
per 100 000 adults in 1990. The average rate of back 9 Gyntelberg F. One year incidence of low back pain among male
residents of Copenhagen aged 4059. Dan Med Bull 1974; 21: 3036.
surgery for 198890 were lowest in the west (113 per 10 Hirsch C, Jonsson B, Lewin T. Low back pain symptoms in a Swedish
100 000 adults) and highest in the south (171 per female population. Clin Orthop 1969; 63: 17116.
100 000 adults). Similar regional trends were also found 11 Hult L. Cervical, dorsal and lumbar spinal syndromes. Acta Orthop
for non-surgical hospital admissions. Scand 1954; 17 (suppl): 1102.
12 Magora A. Investigation of the relation between low back pain and
There are substantial variations between countries in occupation. 2. Work history. Ind Med Surg 1970; 39: 50410.
rates of back surgery. Cherkin and colleagues46 compared 13 Nagi SZ, Riley LE, Newby LG. Social epidemiology of back pain in a
the rates of surgery for back pain in 13 countries and general population. J Chronic Dis 1973; 26: 76979.
Canadian provinces and found that in the USA, the rate 14 Papageorgiou AC, Croft PR, Ferry S, Jayson MIV, Silman AJ.
Estimating the prevalence of low back pain in the general population.
is at least 40% higher than in any other country, and Spine 1995; 20: 188994.
more than five times higher than in Scotland and England 15 Svensson HO, Andersson GBJ, Johansson S, Wilhelmsson C,
(figure 2). Differences in the underlying prevalence of Vedin A. A retrospective study of low back pain in 38- to 64-year-old
back pain are unlikely explanations for the differences in women. Frequency and occurrence and impact on medical services.
Spine 1988; 13: 54852.
surgical rates. More likely explanations are cultural
16 Svensson HO, Andersson GBJ. Low back pain in forty to forty-seven
differences, differences in practice patterns, and the year old men. 1. Frequency of occurrence and impact on medical
availability of health-care providers. services. Scand J Rehab Med 1982; 14: 4753.
17 Valkenburg HA, Haanen HCM. The epidemiology of low back pain.
In: White AA, Gordon SL, eds. Symposium on idiopathic low back
Conclusion pain. St Louis: Mosby, 1982: 922.
Back pain of at least moderate intensity and duration has 18 Walsh K, Cruddas M, Goggon D. Low back pain in eight areas of
an annual incidence in the adult population of 1015%, Britain. J Epidemiol Community Health 1992; 46: 22730.
and a point prevalence of 1530%. The prevalence rises 19 Rossignol M, Suissa S, Abenhaim L. Working disability due to
occupational back pain: three-year follow-up of 2300 compensated
with increasing age up to 65 years, after which age it workers in Quebec. J Occup Med 1988; 30: 50205.
drops off for unknown reasons. The outlook for patients 20 van Doorn TWC. Low back disability among self-employed dentists,
with back pain is generally excellent with 90% or greater veterinarians, physicians and physical therapists in the Netherlands.
Acta Orthop Scand 1995; 66 (suppl 263): 164.
recovery over 3 months. Unfortunately, for individuals
21 Polatin PB, Kinney RK, Gatchel RJ, Lillo E, Mayer TG. Psychiatric
who did not recover within this time the recovery process illness and chronic back pain. The mind and the spinewhich goes
is slow and their demand on the health-care system is first? Spine 1993; 18: 6671.

584 THE LANCET Vol 354 August 14, 1999


22 Spitzer WO, LeBlanc FE, Dupuis M, et al. Scientific approach to the 35 Lancourt J, Kettelhut M. Predicting return to work for lower back
assessment and management of activity-related spinal disorders: a pain patients receiving workers compensation. Spine 1992; 17:
monograph for clinicians. Report of the Quebec Task Force on Spinal 62940.
Disorders. Spine 1987; 12: S1S59. 36 Abenhaim L, Suissa S, Rossignol M. Risk of recurrence of
23 Andersson GBJ, Svensson HO, Oden A. The intensity of work occupational back pain over three year follow-up. Br J Ind Med 1988;
recovery in low back pain. Spine 1983; 8: 88084. 45: 82933.
24 Sander RA, Meyers JE. The relationship of disability to compensation 37 Shekelle PG, Markovich M, Louie R. An epidemiologic study of
status in railroad workers. Spine 1986; 11: 14143. episodes of back care. Spine 1995; 20: 166873.
25 Greenough CG. Recovery from low back pain. 15 year follow-up of 38 Choler U, Larsson R, Nachemson A, Peterson LE. Back pain. Spri
287 injury-related cases. Acta Orthop Scand 1993; 64: (suppl 254): report 188. Stockhom: Spri, 1985.
164. 39 Webster BS, Snook SH. The cost of 1989 workers compensation for
26 Leavitt F. The physical exertion factor in compensable work injuries: a low back pain claims. Spine 1994; 19: 11116.
hidden flaw in previous research. Spine 1992; 17: 30710. 40 Anderson JA. Epidemiological aspects of back pain. J Soc Occup Med
27 Afacan AS. Sickness absence due to back lesions in coal miners. J Soc 1986; 36: 9094.
Occup Med 1982; 32: 2631. 41 Bergquist-Ullman M, Larsson U. Acute low back pain in industry: a
28 Battie MC. The reliability of physical factors as predictors of the controlled prospective study with special reference to therapy and
occurrence of back pain reports. A prospective study within industry. confounding factors. Acta Orthop Scand 1977; 170 (suppl): 1117.
PhD thesis. Goteborg: University of Goteborg, 1989. 42 Biering-Sorensen F. A prospective study of low back pain in a general
29 Battie MC, Bigos SJ, Fisher LD, et al. Anthropometric and clinical population. I. Occurrence, recurrence and aetiology. Scand J Rehabil
measurements as predictors of industrial back pain complaints: a Med 1983; 15: 7179.
prospective study. J Spinal Dis 1990; 3: 195204. 43 Moens GF, Dohogne T, Jaques P, Van Helshoecht P. Back pain and
30 Bigos SJ, Spengler DM, Martin NA, Zeh J, Fisher L, Nachemson A. its correlates among workers in family care. Occup Med 1993; 43:
Back injuries in industry: a retrospective study. III. Employee-related 7884.
factors. Spine 1986; 11: 25256. 44 Koch H. The management of chronic pain in office-based ambulatory
31 Spengler DM, Bigos SJ, Martin NA, Zeh J, Fisher L, Nachemson A. care: national ambulatory medical care survey. Advance data from
Back injuries in industry: a retrospective study. I. Overview and cost vital and health statistics, no 123. National Center for Health Statistics
analysis. Spine 1986; 11: 24145. 1986. DHHS Pub No (PHS) 86-1250. Hyattsville, MD: Public
32 Frymoyer JW, Cats-Baril W. An overview of the incidence and costs of Health Service, 1986.
low back pain. Orthop Clin North Am 1991; 22: 26371. 45 Volinn E, Turczyn KM, Loeser JD. Patterns in low back pain
33 Frymoyer JW, Cats-Baril W. Predictors of low back pain disability. hospitalization: implications for the treatment of low back pain in an
Clin Orthop 1987; 221: 8998. era of health care reform. Clin J Pain 1994; 10: 6470.
34 Burton KA, Tillotson KM, Main CJ, Hollis S. Psychosocial predictors 46 Cherkin DC, Deyo RA, Loeser JD, Bush T, Waddell G. An
of outcome in acute and subchronic low back trouble. Spine 1995; international comparison of back surgery rates. Spine 1994; 19:
20: 72228. 120106.

THE LANCET Vol 354 August 14, 1999 585

You might also like